Data orchestration is the process through which marketing data is analyzed, segmented, and activated at scale, across the entire sales and marketing tech stack.
Let's be honest. Data orchestration is complex, difficult, and well beyond the capabilities of the typical tech stack. That means it's frustrating and discouraging, which is why many companies cut corners, search for silver bullets, implement intent data with unrealistic expectations, and attribute underperformance to other factors.
And yet it is the only way to fully unlock the value of intent data, technographics, and other components of the full marketing data stack. In fact, proper data orchestration is fundamental to account-based marketing (ABM) and other advanced marketing and sales tactics.
Data in the abstract is useless.
Orchestration involves several high-level steps:
So step by step, what's involved in architecting and implementing robust marketing data orchestration?
Sounds simple but too often companies jump on the data bandwagon, connect a bunch of APIs and then follow the same playbooks. What's the goal of data? To support an account-based marketing program? To focus outbound SDRs on specific active opportunities? To help coordinate efforts across target accounts?
Your goals, and the data models that will support them, needn't be mutually exclusive. However, planning needs to anticipate the use cases early to ensure that at a most basic level you have budgeted for the correct data and your tech stack will handle it.
From clear goals you can model the analysis, segmentation and actions that will be required. You don't have to go too deep initially, but need to confirm that conversational marketing, ad management, CRM, sales acceleration, and marketing automation platforms are compatible (along with any others you'll integrate.) Also, in B2B orchestration applications it's critical that you plan for both contact AND account factors.
Even simple marketing data stacks are complex. Multiple identifiers from previously siloed data, typos, contact entered data (forms), rep entered data, imported data, personal and business emails, job changes, etc. all complicate data hygiene in simple data stacks. Add in anonymous first-party data, second-party data, third-party intent data, firmographics, parent/child accounts, sales channel data, and technographics and you've got a challenge.
Effective data orchestration requires effective unification - weaving it together and performing data hygiene functions. It often involves enrichment too, to validate/add emails, phone, mailing addresses, and other factors.
A customer data platform (CDP) may be necessary to achieve this hidden, boring, and critical step.
Marketing automation platforms provide simple toolsets to start with this. For instance, once a contact has visited certain pages, clicked/opened certain emails, submitted certain forms, then you might dynamically define them as marketing qualified.
The analysis engine and data orchestration frameworks will demand less, or more, according to your goals and models. For instance, as soon as three contacts with key job titles appear in the aggregated data, you may well trigger very different sorts of actions depending on the combination of website activity and intent data signals than you would if you only saw two with lower-level titles and less significant types of engagement.
Your analysis engine will need to provide flexibility in creating and modifying models and in running multiple parallel models simultaneously. Increasingly, data orchestration platforms are promoting AI (artificial intelligence) capabilities to suggest enhancements to your models.
That continuous analysis will inform the dynamic segmentation of contacts and accounts which will be used to automate sales and marketing actions, and power sales enablement, at scale. It also supports lead and opportunity scoring, and more advanced propensity to buy modeling.
Segmentation for optimized data orchestration is substantially more complex than the typical contact segmentation based on, for instance, whether they clicked a particular link in the last marketing email. It needs to simultaneously consider account fit and activity, along with contact fit and activity. Additionally, it must weigh factors well beyond 1st party data (engagements with your company's digital footprint.) At a minimum, it should include factors that account for a competitor engagement vector (absolute magnitude and direction, as well as compared to their engagement with your company,) details of job function and seniority in the context of the activity, and insights into buying team activity.
This segmentation - both the dynamic addition to and removal from segments - needs to happen in real-time based on unified data from the entire data stack.
What would your optimal nurturing, demand gen, or sales process call for when a contact (or account) suddenly manifests appropriate levels of interest? You'd potentially do some of the following as examples of numerous possible activities:
Doubtless, your specific sales process would dictate a number of additional, and some different steps. These are simply examples.
Since you've considered connectors, integrations, and API capability (be sure to ask about bi-directional since this is dynamic) earlier, you'll be able to implement any that are important.
Everything that we're doing with data orchestration is to foster meaningful sales engagements. That typically means by sales reps. So while orchestration often focuses on the automatic execution of demand generation and nurturing steps at scale, it's important to keep the end goal in mind.
The people (typically marketing) that have created the vision for the combination of data stack elements and integration of technology will often have the best ability to extract insights from the data and draw reasonable inferences. It's really important to make those readily available for reps, in line with their normal workflows, and to push the dots close together to encourage the incorporation of the insights into their effort.
That's the goal of the sales enablement step of data orchestration which might include plays like:
All routine notifications and internal communications with reps (e.g. contact assignment notifications) should also include some insight into context, what marketing infers might be happening, and why specific recommendations are offered. Proper orchestration will allow reps to incorporate information into their normal process and defined movements. Otherwise, data just sits in a database (often CRM,) confuses reps, and confounds marketers who have committed resources to make it available to sales only to see it lie fallow.
Most of the orchestration steps encounter a barrier in a typical Marketo/HubSpot + Salesforce tech stack. While the process is generally intuitive, the execution is difficult.
Ingestion of multiple data sources and unification is a step that normally requires a CDP or data lake of some sort.
Analysis and segmentation, particularly stitching together contact and account activity, often exceeds the ability of contact-centric marketing automation platforms. You may find, though, that well-unified data can be run through a BI tool to execute some simple executions.
Complete implementation will almost certainly require a purpose-built data orchestration platform.
If you're clear in establishing your goals and diligent in confirming that required platforms can be integrated, then you can be confident in your ability to plan your tech stack and progressively implement increasingly complex movements. And you'll be aware of how far you can proceed without committing to a robust unification engine and flexible data orchestration platform.
ABM is an important example of data orchestration. It involves many of the steps outlined above. However, many of the platforms which are sold expressly for ABM will fail to deliver some of these critical capabilities. For instance, they may add contacts to ad campaigns but prove incapable of important analysis, segmentation, and enablement. Be sure to carefully assess platform capabilities in detail.